Protective Role of Surfactant Protein-D Against Lung Injury and Oxidative Stress Induced by Nitrogen Mustard

Abstract

Nitrogen mustard (NM) is a vesicant known to cause acute pulmonary injury which progresses to fibrosis. Macrophages contribute to both of these pathologies. Surfactant protein (SP)-D is a pulmonary collectin that suppresses lung macrophage activity. Herein, we analyzed the effects of loss of SP-D on NM-induced macrophage activation and lung toxicity. Wild-type (WT) and SP-D^−/− mice were treated intratracheally with PBS or NM (0.08 mg/kg). Bronchoalveolar lavage (BAL) fluid and tissue were collected 14 days later. In WT mice, NM caused an increase in total SP-D levels in BAL; multiple lower molecular weight forms of SP-D were also identified, consistent with lung injury and oxidative stress. Flow cytometric analysis of BAL cells from NM treated WT mice revealed the presence of proinflammatory and anti-inflammatory macrophages. Whereas loss of SP-D had no effect on numbers of these cells, their activation state, as measured by proinflammatory (iNOS, MMP-9), and anti-inflammatory (MR-1, Ym-1) protein expression, was amplified. Loss of SP-D also exacerbated NM-induced oxidative stress and alveolar epithelial injury, as reflected by increases in heme oxygenase-1 expression, and BAL cell and protein content. This was correlated with alterations in pulmonary mechanics. In NM-treated SP-D^−/−, but not WT mice, there was evidence of edema, epithelial hypertrophy and hyperplasia, bronchiectasis, and fibrosis, as well as increases in BAL phospholipid content. These data demonstrate that activated lung macrophages play a role in NM-induced lung injury and oxidative stress. Elucidating mechanisms regulating macrophage activity may be important in developing therapeutics to treat mustard-induced lung injury.

Nitrogen mustard (NM) is a cytotoxic vesicant known to target the lung. Both acute and long-term effects have been described following NM exposure, including inflammation and disruption of the alveolar epithelial barrier, hemorrhage, edema, intra-alveolar septal thickening, bronchiolitis, and pulmonary fibrosis (Sunil et al., 2014; Venosa et al., 2016). Toxicity results from alkylation and cross-linking of a variety of cellular components including nucleic acids, proteins, and lipids, leading to impairment of cellular functioning and cell death; this is associated with oxidative and nitrosative stress (Korkmaz et al., 2006). Inflammatory macrophages accumulating in the lung in response to NM-induced injury have been implicated in its pathogenic effects by promoting oxidative and nitrosative stress and releasing proinflammatory, cytotoxic, and profibrotic mediators (Malaviya et al., 2016a). Evidence suggests that the diverse activities of macrophages are mediated by distinct subpopulations, broadly classified as cytotoxic/proinflammatory M1 and anti-inflammatory/wound repair M2 macrophages, which develop in response to mediators they encounter in the tissue microenvironment (Arora et al., 2018; Laskin et al., 2011). Whereas overactivation of M1 macrophages promotes tissue injury and oxidative stress, excessive activity of M2 macrophages contributes to fibrosis.

Macrophage activation in the lung is controlled, in part, by surfactant protein (SP)-D, a pulmonary collectin synthesized by Type II alveolar epithelial cells. Under homeostatic conditions, SP-D functions as an anti-inflammatory protein, suppressing macrophage proinflammatory responses. However, following lung injury, SP-D can undergo structural alterations resulting in the generation of dimers and trimers which promote proinflammatory activity (Atochina-Vasserman, 2012). Reactive nitrogen species (RNS), generated in part by inflammatory macrophages, are thought to play a role in disruption of the SP-D molecule changing its activity. Thus, in models of lung injury induced by ozone, bleomycin, or radiation, SP-D is S-nitrosylated resulting in exacerbation of inflammation and toxicity (Groves et al., 2012; Guo et al., 2016; Malaviya et al., 2015a). Analogous increases in inflammation and tissue injury in response to endotoxin, ozone or bleomycin have been observed in mice lacking functional SP-D (Casey et al., 2005; Groves et al., 2012; King et al., 2011). The role of SP-D in regulating macrophage activity in the lung following NM exposure is unknown and this represents the focus of the present studies. Our findings that lung macrophage activation is amplified in mice lacking SP-D, and that this correlates with increased sensitivity to the toxic effects of NM provide support for the idea that activated macrophages and mediators they release contribute to acute lung injury. Identification of specific pathways regulating the activity of lung macrophages may be useful in the development of efficacious approaches to mitigating morbidity and mortality following exposure to mustards.

MATERIALS AND METHODS

Animals and exposures

Male and female specific pathogen-free C57Bl6/J mice (21–27 weeks; 17–30 g) were obtained from The Jackson Laboratories (Bar Harbor, Maine). SP-D^−/− mice, generated by targeted gene inactivation, followed by back crossing several generations into C57Bl/6 mice (Atochina et al., 2004), were bred at the Rutgers University animal facility. Animals were housed in filter-top microisolation cages and maintained on food and water ad libitum. All animals received humane care in compliance with the institution’s guidelines, as outlined in the Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health. To ensure delivery of NM to the lower lung, a major target in humans, mice were exposed intratracheally. Animals were anesthetized using isoflurane (Henry Schein, Melville, New York), and then placed on a titling rodent work stand (Hallowell EMC, Pittsfield, Massachusetts) in a supine position and restrained using an incisor loop. The tongue was extruded using a cotton tip applicator and the larynx visualized by a hemi-sectioned 3 mm speculum attached to an operating head of an otoscope (Welch Allyn, Skaneateles Falls, New York). NM (mechlorethamine hydrochloride, Sigma-Aldrich, St. Louis, Missouri) in PBS was administered intratracheally via Clay Adams Intramedic PE-60 (ID 0.76 mm, OD 1.22 mm) polyethylene tubing (Becton, Dickinson and Company, Franklin Lakes, New Jersey) attached to a 27^1/2-gauge hypodermic needle (0.4 × 13 mm). The tubing was advanced approximately 10 mm past the epiglottis and 0.05 ml of sterile PBS or NM (0.08 mg/kg) was instilled into the trachea. The tubing and speculum were withdrawn immediately after instillation. Animals were then removed from the work stand and maintained in a vertical position until normal respiration was observed (less than 1 min). All intratracheal instillations were performed by David Reimer, DVM, Rutgers University Comparative Medicine Resources. In pilot time course studies (Sunil et al., 2017), we found that the most pronounced effects of NM were observed 14 days after treatment of WT mice. We therefore selected this post-exposure time point for comparative analysis of WT and SP-D^−/− mice.

BAL, lung cells, and tissue collection

Animals were euthanized 14 days after exposure by intraperitoneal injection of Sleepaway^® (sodium pentobarbital, 1.3 μg/kg; Fort Dodge Animal Health, Fort Dodge, Iowa). BAL was collected by slowly instilling and withdrawing 1 ml of ice-cold PBS into the lung three times through a 20-gauge cannula inserted into the trachea. BAL fluid was centrifuged (300 × g, 8 min), supernatants collected, aliquoted, and stored at −80°C until analysis. Cell pellets were resuspended in 250 μl PBS and viable cells counted on a hemocytometer using trypan blue dye exclusion. For differential analysis, cytospin preparations (1 × 10⁵ BAL cells/slide) were fixed in methanol and stained with Giemsa (Labchem Inc., Pittsburgh, Pennsylvania). A total of 300 cells were counted by light microscopy.

To isolate cells for flow cytometry, the lung was perfused in situ via the portal vein with 5 ml of warm (37°C) perfusion medium (25 mM HEPES, 0.5 mM EGTA, 4.4 mM NaHCO₃ in HBSS, pH 7.3), followed by perfusion with Ca⁺²/Mg⁺²-free HBSS (22 mM HEPES, 4.2 mM NaHCO₃, pH 7.3) at a rate of 10 ml/min. BAL was collected as described above. The lung was then removed and instilled five times with 1 ml HBSS (37°C), while gently massaging the tissue. Lavage fluid was centrifuged (300 × g, 8 min, 4°C), the cell pellet resuspended in 1 ml PBS and combined with the first BAL lavage cell suspension. Cells were washed three times with HBSS-2% FBS and viable cells counted by trypan blue dye exclusion.

Separate groups of mice were used for tissue collection. The lung was inflated via the trachea with PBS containing 3% paraformaldehyde. After 4 h on ice, the tissue was transferred to 50% ethanol. Histological sections (4 μm) were prepared and stained with hematoxylin and eosin or with Mason’s trichrome stain. Images were acquired at high resolution using an Olympus VS120 Virtual Microscopy System, scanned and viewed using OlyVIA version 2.6 software (Center Valley, Pennsylvania). The extent of inflammation, including macrophage and neutrophil accumulation, alterations in alveolar epithelial barriers, fibrin deposition, edema and fibrosis, were assessed blindly by a board certified veterinary pathologist (Michael Goedken, DVM, PhD), and scored on a scale of 0–4, with 0 indicating no change relative to control; 1 = minimal/very few or small lesions; 2 = slight/few/small lesions; 3 = moderate/moderate size lesions; 4 = large/marked lesions/many changes.

Immunohistochemistry

Tissue sections (4 μm) were deparaffinized with xylene (4 min, × 2), followed by decreasing concentrations of ethanol (100%–50%) and then water. After antigen retrieval using citrate buffer (10.2 mM sodium citrate, 0.05% Tween 20, pH 6.0) and quenching of endogenous peroxidase with 3% H₂O₂ for 10–30 min, sections were incubated for 2 h at room temperature with 5–20% goat serum to block nonspecific binding. This was followed by overnight incubation at 4°C with rabbit IgG or rabbit polyclonal anti-HO-1 (1:250; Enzo Life Sciences, Farmingdale, New York), anti-Ym-1 (1:75; Abcam, Cambridge, Massachusetts), anti-iNOS (1:500; Abcam), anti-COX-2 (1:800; Abcam), anti-MMP-9 (1:50, Santa Cruz Biotechnology, Dallas, Texas), anti-pro-SP-C (1:750, EMD Millipore, Billerica, Massachusetts), or anti-mannose receptor (MR)-1 (1:400; Abcam), antibodies. Sections were then incubated with biotinylated secondary antibody (Vector Labs, Burlingame, California) for 30 min at room temperature. Binding was visualized using a Peroxidase Substrate Kit DAB (Vector Labs). Macrophages and Type II cells were enumerated by light microscopy in all 5 lung lobes (15 fields/lobe; 3–4 mice/treatment group); cells were assigned a staining intensity score of 0–3, with 0 = no staining, 1 = light staining, 2 = medium staining and 3 = dark staining. To account for regional differences seen in response to NM, we counted the number of positively stained cells in all 5 lung lobes (15 fields/lobe).

Flow cytometry

Cells were resuspended in 100 μl of staining buffer (PBS, 2% FCS and 0.02% sodium azide) and incubated for 10 min at 4°C with anti-mouse CD16/32 (1:100; Biolegend, San Diego, California) to block nonspecific binding, followed by FITC-conjugated anti-mouse CD11b (1:200; Biolegend), PE-conjugated anti-mouse Ly6C (1:200; Biolegend), PE/Cy7-conjugated anti-mouse F4/80 (1:200; Biolegend), AF 700-conjugated anti-mouse CD11c (1:200, Biolegend), and AF 647-conjugated anti-mouse Ly6G (1:200; Biolegend) antibodies for 30 min, and then with eFluor 780-conjugated fixable viability dye (1:1000; eBioscience, San Diego, California) for an additional 30 min at 4°C. Cells were washed twice with staining buffer, fixed with 3% paraformaldehyde, and analyzed on a Gallios flow cytometer (Beckman Coulter, Brea, California). Data were analyzed using Beckman Coulter Kaluza version 1.2 software. Lung cell populations were characterized as described previously (Sunil et al., 2015). To obtain cell numbers for the different subpopulations identified by flow cytometry, the percentage of positively stained cells was multiplied by the total number of BAL cells recovered. Representative examples of our gating strategy and cell numbers are shown in Supplementary Figure 1.

Measurement of BAL protein, phospholipids, and SP-D content

Total protein was quantified in cell-free BAL using a BCA Protein Assay kit (Pierce Biotechnologies Inc., Rockford, Illinois) with bovine serum albumin as the standard. Samples (25 μl) from 8 to 9 mice/treatment group were analyzed in triplicate at 560 nm on a Vmax MAXline^TM microplate reader (Molecular Devices, Sunnyvale, California). For phospholipid analysis, BAL was centrifuged at 20 000 × g (1 h, 4°C) and pellets, containing large aggregate fractions, analyzed spectrophotometrically as reported previously (Atochina et al., 2004). To analyze the effects of NM on monomeric SP-D (43 kDa), small aggregate BAL fractions, prepared as described previously (Atochina et al., 2004), were separated on 4–12% Bis-Tris gradient reducing/denaturing gels (Invitrogen, San Diego, California). For analysis of multimeric forms of SP-D, native gel electrophoresis was performed as described in our earlier studies (Guo et al., 2008). Proteins were transferred to PVDF membranes. Non-specific binding was blocked by incubating the blots with 10% milk in T-TBS (0.5% Tween 20 in Tris-buffered saline) for 45 min at room temperature. Blots were then incubated overnight at 4°C with rabbit polyclonal anti-SP-D antibody DU117 (1:20 000; a gift from Amy Pastva, Duke University, North Carolina). Blots were washed and incubated for 1 h at room temperature with HRP-conjugated secondary antibody (1:5000, 1% non-fat milk in T-TBS; Bio-Rad, Hercules, California). Bands were visualized using an ECL Prime detection system (GE Health Care, Piscataway, New Jersey).

Measurement of pulmonary mechanics

Mice were anesthetized with ketamine (80 mg/kg) and xylazine (10 mg/kg). After 5 min, a tracheotomy was performed using an 18-gauge cannula and the animals connected to a SciReq flexiVent (Montreal, Canada) to measure respiratory mechanics, at positive end-expiratory pressure (PEEPs) ranging from 0 to 9 cm H₂O. We specifically selected this analysis as it reflects alterations in the lower lung, the primary focus of our studies (D′Angelo et al., 1992; Smith et al., 2013). Data from the impedance spectra were fit to a constant phase model, allowing for the calculation of frequency independent central airway resistance (R_aw), and the coefficients of tissue damping (G) and tissue elastance (H) in the tissue compartment (Suki et al., 1991). Data were analyzed using flexiVent software version 5.2.

Statistical analysis

All experiments were repeated at least 3 times. Data were analyzed using 2-way or 3-way ANOVA followed by post hoc analysis using the Mann–Whitney rank sum test; a p value ≤ .05 was considered statistically significant.

RESULTS

Effects of NM on SP-D Structure

Initially, we analyzed the effects of NM administration to mice on intact monomeric SP-D using denaturing gel electrophoresis. Relatively low levels of SP-D were detected in BAL from WT mice treated with PBS (Figure 1, top panel). An increase in monomeric SP-D levels was observed following NM administration. To assess NM-induced alterations in SP-D structure, BAL samples were analyzed by native gel electrophoresis. The secreted form of SP-D is known to assemble into large cruciform dodecamers composed of a tetramer of trimers that do not readily migrate in native gels (Guo et al., 2008). Consistent with this observation, only intact SP-D dodecamers were detected in BAL from WT mice treated with PBS control (Figure 1, bottom panel). Following NM administration, multiple lower molecular weight forms of SP-D were observed, including trimers and dimers, indicative of lung injury. As expected, SP-D was not detectable in BAL from PBS- or NM-treated SP-D^−/− mice, although some non-specific binding was evident.

Figure 1.

Effects of NM on BAL SP-D levels and structure. BAL was collected 14 days after treatment of WT and SP-D^−/− mice with PBS or NM and analyzed by Western blotting as described in Materials and Methods section. Upper panel: SP-D protein analyzed in denaturing gels. Lower panel: SP-D protein analyzed in native gel electrophoresis. BAL samples containing equivalent amounts of SP-D were subjected to native electrophoresis to determine multimeric structure of SP-D. Each lane represents an individual animal.

Effects of Loss of SP-D on the Response of Lung Macrophages to NM

To characterize the phenotype of macrophages responding to NM and the impact of loss of SP-D, we used techniques in flow cytometry (Sunil et al., 2015). In these experiments, cells were first analyzed for expression of CD11b, a β2-integrin expressed on infiltrating myeloid cells. This was followed by analysis of the granulocytic marker, Ly6G, the macrophage activation marker, Ly6C, and the mature lung macrophage markers, F4/80 and CD11c. Treatment of WT mice with NM resulted in a significant increase in infiltrating CD11b⁺Ly6G⁻ monocytic cells in the lung, with no effects on CD11b⁺Ly6G⁺ granulocytic cells (Figure 2, top panels). Loss of SP-D had no significant effects on the monocytic cells or their response to NM. In SP-D^−/− mice treated with PBS control, increased numbers of CD11b⁺Ly6G⁺ granulocytic cells were observed in the lung, relative to WT mice; however, these cells did not change after NM (Figure 2, top right panel). Mature inflammatory macrophages (CD11b⁺Ly6G⁻F4/80⁺CD11c⁺) were also identified in the lung; these consisted of proinflammatory Ly6C^hi and anti-inflammatory Ly6C^lo subpopulations (Figure 2, middle panels). Increases in both of these inflammatory macrophage subpopulations were observed in the lung after NM; loss of SP-D had no effect on these cells.

Figure 2.

Effects of loss of SP-D on NM-induced inflammatory cell accumulation in the lung. Cells, collected by lavage and massage 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with antibodies to CD11b, Ly6G, Ly6C, F4/80, and CD11c or appropriate isotypic controls, and analyzed by flow cytometry. Bars, mean ± SE (n = 8–11 mice/treatment group). ^aSignificantly different (p < .05) from PBS treated mice. ^bSignificantly different (p < .05) from WT mice. M-MDSC, monocytic-myeloid derived suppressor cell; G-MDSC, granulocytic-myeloid derived suppressor cell; ND, not detected.

Further analysis of the CD11b⁺Ly6G⁺ granulocytic subpopulation revealed that they also expressed high levels of Ly6C, a characteristic of myeloid-derived suppressor cells (MDSC) (Kong et al., 2013). These cells consisted of F4/80⁺ monocytic (M) and F4/80⁻ granulocytic (G) subpopulations; the majority were M-MDSC (Figure 2, lower panels). Greater numbers of M-MDSC and G-MDSC were observed in lungs of PBS treated SP-D^−/− mice, when compared with WT mice. NM had no effect on MDSC in either WT or SP-D^−/− mice.

Consistent with previous studies (Matthews et al., 2007), resident alveolar macrophages were identified as CD11b⁻CD11c⁺F4/80⁺Ly6G⁻Ly6C^lo. Increased numbers of these cells were observed after NM exposure; this response was significantly greater in SP-D^−/− mice relative to WT mice (Table 1).

Table 1.

Effects of Loss of SP-D on the Response of Resident Alveolar Macrophages (CD11b⁻CD11c⁺F4/80⁺Ly6G⁻Ly6C^lo) to NM

	WT	SP-D−/−
PBS	1299.7 ± 161.8	1671.1 ± 226.1
NM	2516.7 ± 308.3a	4485.1 ± 202.3a,b

BAL cells, collected by lavage and massage 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with antibodies to CD11b, CD11c, Ly6G, Ly6C, and F4/80 or appropriate isotypic controls, and analyzed by flow cytometry. Data are mean of cells ± SE (n = 8–11 mice/treatment group).

^aSignificantly different (p < .05) from PBS.

^bSignificantly different (p < .05) from WT.

We next determined if macrophages accumulating in the lung following NM exposure were functionally activated, by analyzing expression of proinflammatory M1 (iNOS, MMP-9, COX-2) and anti-inflammatory M2 (Ym-1, MR-1) macrophage activation markers in histological sections (Laskin et al., 2011). Treatment of WT mice with NM resulted in increased numbers of lung macrophages staining positively for iNOS and MMP-9 (Figure 3 and Table 2). The majority of these cells had a staining intensity of 2 (medium staining). Loss of SP-D resulted in a significant increase in macrophages staining for iNOS and MMP-9 in response to NM; macrophages staining for MMP-9 were also increased in size (Figure 3). Increased expression of iNOS was also observed in Type II cells after NM (Figure 3), suggesting an activated phenotype. This is supported by findings that COX-2 and pro-SPC were also upregulated in Type II cells following NM administration (Figure 4 and Table 2). Loss of SP-D was associated with a significant increase in NM-induced COX-2 expression in Type II cells with no major effect on pro-SP-C expression (Figure 4 and Table 2). In contrast, macrophage expression of COX-2 was relatively low in both WT and SP-D^−/− mice, even after NM administration (Figure 4 and Table 2). Macrophages did, however, express pro-SP-C after NM; loss of SP-D had no effect on this response. We also analyzed the activation profile of M2 macrophages. In PBS-treated WT mice, macrophage staining for MR-1, but not Ym-1 was noted (Figure 5 and Table 2). NM treatment of WT mice was associated with increased numbers of macrophages staining for Ym-1 and MR-1 (Figure 5 and Table 2). Loss of SP-D resulted in a marked increase in both intensity and number of macrophages staining for Ym-1 and MR-1, most notably in enlarged foamy macrophages (Figure 5 and Table 2). No staining for Ym-1 or MR-1 was observed in Type II cells.

Figure 3.

Effects of loss of SP-D on NM-induced iNOS and MMP-9 expression. Histological sections, prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with antibody to iNOS (upper panels) or MMP-9 (lower panels). Binding was visualized using a DAB peroxidase substrate kit. One representative section (staining intensity, 2) from 3 to 4 mice/treatment group is shown. Magnification, 60×; arrows, macrophages; arrowheads, Type II cells.

Table 2.

Semi-Quantitation of Immunohistochemistry

	WT				SP-D−/−
	Staining intensity				Staining intensity
Lung Macrophages		1	2	3	1	2	3
iNOS	PBS	1.2 ± 0.1	0.4 ± 0.1	0.0 ± 0 .0	0.7 ± 0.1	0.0 ± 0.0	0.0 ± 0.0
	NM	4.4 ± 0.7a	10.1 ± 3.2a	0.7 ± 0.4a	13.2 ± 0.3a, b	18.7 ± 1.3a, b	0.9 ± 0.2a, b
MMP-9	PBS	10.1 ± 0.8	1.6 ± 0.3	0.0 ± 0.0	7.5 ± 1.0	12.6 ± 3.7	0.0 ± 0.0
	NM	18.3 ± 2.3a	37.7± 5.2a	0.0 ± 0.0	6.2 ± 1.6a	128.2 ± 3.7a, b	1.3 ± 0.6
COX-2	PBS	0.4 ± 0.3	0.0 ± 0.0	0.0 ± 0.0	0.6 ± 0.2	0.2 ± 0.1	0.0 ± 0.0
	NM	2.5 ± 0.5a	0.3 ± 0.2a	0.0 ± 0.0	2.8 ± 1.0a	1.4 ± 0.4a, b	0.0 ± 0.0
Pro-SPC	PBS	5.7 ± 1.5	8.0 ± 3.3	0.0 ± 0.0	8.5 ± 2.1	1.7 ± 0.7	0.0 ± 0.0
	NM	16.7 ± 3.1a	17.2 ± 3.6a	0.0 ± 0.0	14.1 ± 1.8a	22.8 ± 4.3a	0.0 ± 0.0
YM-1	PBS	7.3 ± 1.0	3.4 ± 1.1	0.0 ± 0.0	6.7 ± 1.0	14.4 ± 2.5	1.5 ± 0.6
	NM	21.5 ± 3.4a	40.6 ± 3.7a	0.0 ± 0.0	20.0 ± 5.7a	52.4 ± 6.2a	11.8 ± 2.9a, b
MR-1	PBS	3.4 ± 0.0	31.0 ± 0.0	0.0 ± 0.0	16.3 ± 2.6	17.1 ± 2.9	0.6 ± 0.3
	NM	7.9 ± 1.7	34.7 ± 85.5	23.8 ± 3.8a	27.2 ± 6.8b	79.9 ± 11.5a, b	49.8 ± 7.8a, b
HO-1	PBS	2.7 ± 0.6	0.0 ± 0.0	0.0 ± 0.0	7.0 ± 1.2	10.4 ± 1.7	0.0 ± 0.0
	NM	10.6 ± 1.1a	23.4 ± 3.3a	0.2 ± 0.1	4.0 ± 1.2b	81.0 ± 18.2a, b	0.7 ± 0.4a
Type II Cells
COX-2	PBS	5.3 ± 1.1	6.9 ± 1.3	0.0 ± 0.0	7.5 ± 0.7	11.3 ± 0.9	0.0 ± 0.0
	NM	14.7 ± 1.8a	47.7 ± 8.1a	1.1 ± 0.1a	9.5 ± 1.6b	135.7 ± 9.9a, b	4.7 ± 1.9a
Pro-SP-C	PBS	25.2 ± 4.7	22.1 ± 6.2	0.0 ± 0.0	10.7 ± 2.1	5.2 ± 2.6	0.0 ± 0.0
	NM	18.3 ± 2.5	129.8 ± 5.9a	1.7 ± 1.2a	28.9 ± 3.4a, b	136.1 ± 10.8a	1.2 ± 0.5a
HO-1	PBS	5.7 ± 0.9	0.0 ± 0.0	0.0 ± 0.0	14.7 ± 1.6	7.4 ± 1.8	0.0 ± 0.0
	NM	19.8 ± 2.7a	45.9 ± 6.4a	0.0 ± 0.0	16.1 ± 6.3	93.3 ± 11.8a, b	2.1 ± 1.2a, b

Lung sections were prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM. Numbers of lung macrophages staining positive for inducible nitric oxide synthase (iNOS), matrix metallo-proteinase (MMP)-9, cyclooxygenase (COX)-2, pro-SP-C, Ym-1, mannose receptor (MR)-1 and hemeoxygenase (HO)-1, and Type II cells staining positive for COX-2, pro-SP-C and HO-1, were enumerated in all five lung lobes [15 fields (40×)/lobe] and assigned a staining intensity score of 0–3 (0, no staining; 1, light staining; 2, medium staining; 3, dark staining). Values are the mean ± SE of 3–4 mice/treatment group. Data were analyzed by ANOVA.

^aSignificantly different (p < .05) from PBS.

^bSignificantly different (p < .05) from WT.

Figure 4.

Effects of loss of SP-D on NM-induced COX-2 and Pro-SP-C expression. Histological sections, prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with antibody to COX-2 (upper panels) or Pro-SP-C (lower panels). Binding was visualized using a DAB peroxidase substrate kit. One representative section (staining intensity, 2) from 3 to 4 mice/treatment group is shown. Magnification, 60×; arrows, macrophages; arrowheads, Type II cells.

Figure 5.

Effects of loss of SP-D on NM-induced Ym-1 and MR-1 expression. Histological sections, prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with antibody to Ym-1 (upper panels) or MR-1 (lower panels). Binding was visualized using a DAB peroxidase substrate kit. One representative section (Ym-1, staining intensity, 2; MR-1, staining intensity, 3) from 3 to 4 mice/treatment group is shown. Magnification, 60×; arrows, macrophages; arrowheads, Type II cells.

Loss of SP-D Promotes NM-Induced Lung Injury, Inflammation and Oxidative Stress

In further studies, we determined if heightened activation of M1 and M2 macrophages in lungs of NM-treated SP-D^−/− mice was associated with exacerbation of histopathology and oxidative stress. In WT mice, NM-induced histopathological changes in the lung including perivascular inflammation and edema; these were more pronounced in SP-D^−/− mice (Figure 6 and Table 3). Additionally, in SP-D^−/− mice, but not in WT mice, NM caused hyperplasia of alveolar-bronchiolar epithelial and goblet cells, bronchiectasis, emphysema, and fibrotic changes, including interstitial fibroplasia and collagen deposition. Macrophages accumulating in the lungs of SP-D^−/− mice in response to NM were also enlarged and highly vacuolated, relative to macrophages in lungs of WT mice (Figure 6 and Table 3). Increases in enlarged vacuolated macrophages were also noted in PBS treated SP-D^−/−, along with perivascular inflammation, edema, and air sac enlargement.

Figure 6.

Effects of loss of SP-D on NM-induced structural alterations in the lung. Lung sections, prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with H&E (upper panels) or Mason’s trichrome stain (lower panels). One representative section from 3 to 4 mice/treatment group is shown. Upper panel magnification, 20×; insets, alveolar macrophages, 40×; a, alveolar macrophages; b, peri-vascular inflammation; c, peri-vascular edema; d, hyperplasia of bronchiolar epithelium; e, hyperplasia of alveolar epithelium; Lower panel magnification, 13.2×; col, collagen deposits.

Table 3.

Quantitation of Histopathological Changes in the Respiratory Tract

	PBS		NM
	WT	SP-D−/−	WT	SP-D−/−
Bronchioles
Hypertrophy/hyperplasia bronchiolar epithelium	0	0	0	2.3 ± 0.3b
Hyperplasia bronchiolar alveolar	0	0	0	2.7 ± 0.3b
Metaplasia	0	0	0	0.3 ± 0.3
Bronchiectasis	0	0	0	3.3 ± 0.7b
Goblet cell hyperplasia	0	0	0	3.0 ± 0.0b
Parenchyma
Foamy macrophages (absolute number mainly in alveolar space)	0	0.8 ± 0.3	0	2.3 ± 0.7b
Foamy macrophages (size and number of vacuoles)	0	0.8 ± 0.5	0	2.7 ± 0.3a,b
Perivascular inflammation	0	0.5 ± 0.5	0.7 ± 0.3	1.7 ± 0.3
Peribronchiolar macrophages	0	0	0	0
Bronchiolization	0	0	0	0.3 ± 0.3
Edema	0	1.0 ± 0.0	0.3 ± 0.3	1.0 ± 0.0
Air sac enlargement	0	0.5 ± 0.3	0	0.3 ± 0.3
Fibrotic changes (trichrome)
Interstitial fibroplasia	0	1.0 ± 0.4	0	1.7 ± 0.7b
Fibrosis	0	0	0	1.3 ± 0.3a, b
Emphysema	0	0	0	2.0 ± 0.0a, b

Lung sections, prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM were stained with H&E or Mason’s trichrome stain and scored for pathological changes on a scale of 0–4 with 0, no significant lesions/no pathological findings; 1, minimal/very few or small lesions; 2, slight/few/small lesions; 3, moderate/moderate size; 4, large/marked/many changes. Data are mean ± SE (n = 3–4 mice/treatment group).

^aSignificantly different (p < .05) from PBS.

^bSignificantly different (p < .05) from WT.

NM treatment of WT mice resulted in increases in BAL protein and cell content, markers of alveolar epithelial barrier dysfunction; these effects were greater in SP-D^−/− mice (Figure 7). Moreover, the percentage of neutrophils in BAL from SP-D^−/− mice was 2-fold greater than in WT mice (Supplementary Table 1) following NM exposure. Total BAL phospholipids were also increased after NM exposure, but this was only observed in SP-D^−/− mice (Figure 7).

Figure 7.

Effects of loss of SP-D on NM-induced increases in BAL cell number, protein and total phospholipid content. BAL was collected 14 days after treatment of WT and SP-D^−/− mice with PBS or NM. Upper panel: Viable cells were enumerated by trypan blue dye exclusion. Bars, mean ± SE (n = 11–14 mice/treatment group). Middle panel: Cell-free supernatants were analyzed in triplicate for protein using a BCA protein assay kit. Bars, mean ± SE (n = 11–14 mice/treatment group). Lower panel: Total phospholipid content was determined as described in Materials and Methods. Bars, mean ± SE (n = 3–9 mice/treatment group). ^aSignificantly different (p < .05) from PBS; ^bsignificantly different (p < .05) from WT.

We next examined the effects of loss of SP-D on NM-induced expression of HO-1, a marker of oxidative stress (Raval et al., 2010; Wu et al., 2012). Relatively low numbers of HO-1⁺ macrophages were observed in lungs of PBS-treated WT and SP-D^−/− mice (Figure 8 and Table 2). Treatment of mice with NM resulted in increases in HO-1⁺ macrophages in both genotypes; loss of SP-D increased this response (Figure 8). Whereas in WT mice, HO-1⁺ macrophages were relatively small, in SP-D^−/− mice they were enlarged and highly vacuolated. HO-1 expression was also increased in Type II cells after NM administration; loss of SP-D augmented this response (Table 2 and Figure 8).

Figure 8.

Effects of loss of SP-D on NM-induced HO-1 expression. Histological sections, prepared 14 days after treatment of WT and SP-D^−/− mice with PBS or NM, were stained with antibody to HO-1. Binding was visualized using a DAB peroxidase substrate kit. One representative section (staining intensity, 2) from 3 to 4 mice/treatment group is shown. Magnification, 60×; arrows, macrophages; arrowheads, Type II cells.

Effects of Loss of SP-D on NM-Induced Alterations in Pulmonary Mechanics

Following NM administration, total lung resistance increased in both WT and SP-D^−/− mice, a response that was not altered by increasing PEEP, indicating that it was not due to a reversible loss of respiratory units (Figure 9). In contrast, NM had no significant effect on central airway resistance. Differences in the effects of NM on total lung resistance and central airway resistance suggest reduced airflow function in the lung parenchyma. This is supported by our findings that tissue damping increased in both WT and SP-D^−/− mice following NM administration. Tissue elastance was also increased after NM administration, however, the response of SP-D^−/− mice was reduced, when compared with WT mice; a similar reduced response was observed with respect to tissue resistance. Examination of PV loops across PEEPs (0−9 cm H₂O) revealed that lung volume was reduced in SP-D^−/− mice relative to WT mice (Figure 9 and data not shown). NM-induced loss of lung volume was blunted in SP-D^−/− mice, when compared with WT mice at PEEP 3 and PEEP 6 cm H₂O.

Figure 9.

Effects of loss of SP-D on NM-induced alterations in lung function. Upper panels: Total lung resistance, central airway resistance, tissue damping, and tissue elastance at PEEPs ranging from 0 to 9 cm H₂O. Lower panels: Pressure/volume loops at PEEP 3 and PEEP 6 cm H₂O over an inflation/deflation cycle. Measurements were made in triplicate, 14 days after treatment of WT and SP-D^−/− mice with PBS or NM. Data, mean ± SE (n = 3–7 mice/group). For each lung parameter, significant differences between the lines were determined by a 3-way ANOVA comparison of treatment, strain and PEEP. ^aOverall line is significantly different (p < .05) from PBS; ^boverall line is significantly different (p < .05) from WT.

DISCUSSION

Lung injury and fibrosis induced by NM are characterized by a persistent macrophage dominant inflammatory response. In previous studies, we demonstrated that these cells consist of proinflammatory/cytotoxic M1 macrophages and anti-inflammatory/wound repair M2 macrophages, which sequentially accumulate in the lung following NM exposure (Venosa et al., 2016). Findings that the appearance of M1 macrophages in the lung correlates with acute injury and M2 macrophages with fibrosis, suggest that they contribute to these pathologic responses to NM, activities consistent with previous reports in other pulmonary disease models (Jiang et al., 2016; Kawasaki, 2015; Vlahos et al., 2014). The present studies demonstrate that macrophage activation in the lung following NM exposure is controlled, in part, by the pulmonary collectin, SP-D. Thus, in the absence of SP-D, we observed heightened M1 and M2 macrophage activation. The fact that this correlated with exacerbation of NM-induced lung injury and fibrogenesis provides support for the idea that both macrophage subtypes play a role in these pathologic responses.

Increases in SP-D have been reported to be an indicator of pulmonary inflammation (Atochina-Vasserman, 2012; Atochina-Vasserman et al., 2010; Matalon et al., 2009). Consistent with this activity, we found that NM-induced inflammation was associated with an increase in total SP-D levels in the lung. We also found that SP-D was structurally altered after NM exposure, a response linked to lung injury (Guo et al., 2008; Malaviya et al., 2015b). Thus, large SP-D dodecamers disassembled resulting in the formation of small multimers, including trimers and dimers. Earlier studies demonstrated that, in contrast to SP-D dodecamers, which suppress inflammation by blocking macrophage NF-κB, small SP-D multimers promote inflammation by activating NF-κB (Atochina-Vasserman, 2012; Guo et al., 2008). These findings suggest a potential pathway regulating macrophage activation in the lung following NM exposure. Evidence suggests that structural modifications of SP-D are due to iNOS-derived RNS (Atochina-Vasserman, 2012; Malaviya et al., 2015a). Our findings that iNOS is up-regulated in macrophages, as well as Type II cells following NM exposure suggest that both cell types contribute to structural and functional alterations in SP-D.

Flow cytometric analysis of lung cells from WT mice revealed the presence of multiple subpopulations of CD11b⁺ infiltrating myeloid cells, including mature (F4/80⁺CD11c⁺) pro- and anti-inflammatory macrophages, phenotypically identified by high and low expression levels, respectively, of Ly6C. In accord with our findings in rats (Venosa et al., 2016), greater numbers of both of these macrophage subpopulations were observed in the lung after NM exposure. Surprisingly, loss of SP-D had no effect on numbers of these cells. This prompted us to assess whether the macrophages were functionally altered by analyzing expression of M1 and M2 activation markers. Following NM exposure, macrophages accumulating in the lung expressed iNOS and MMP-9, enzymes important in M1 macrophage-induced cytotoxicity (Cunha et al., 2016). Loss of SP-D was associated with a marked up-regulation of macrophage iNOS and MMP-9 expression, suggesting M1 macrophage proinflammatory activity is increased in the absence of SP-D. These findings are consistent with the counter-regulatory role of SP-D in inflammatory lung macrophages (Groves et al., 2012; Guo et al., 2008). In contrast, loss of SP-D had no effect or caused a small increase in macrophage COX-2 expression in response to NM. These data indicate that there are multiple subpopulations of proinflammatory macrophages that are regulated by distinct mechanisms.

NM administration to WT mice was also associated with increases in lung macrophage staining for the M2 macrophage activation proteins, Ym-1 and MR-1. Loss of SP-D amplified this response; thus, macrophage expression of Ym-1 and MR-1 was greater, and numbers of enlarged foamy macrophages expressing these proteins increased. These results were unexpected, as SP-D is generally thought to be a specific suppressor of proinflammatory M1 macrophages (Gardai et al., 2003; Janssen et al., 2008). Increases in activated M2 macrophages in NM-treated SP-D^−/− mice may be due to greater numbers of M1 macrophages that are available for phenotypic switching. Alternatively, it may be that SP-D regulates an early signaling pathway in macrophages that controls both M1 and M2 macrophage activation. Further studies are required to analyze these possibilities.

Alveolar epithelial Type II cells play a key role in the synthesis of surfactant proteins, and in the maintenance of alveolar epithelial barrier functioning (Finkelstein, 1990). Accumulating evidence suggests that they are also active participants in lung inflammatory responses (Fehrenbach, 2001). The present studies demonstrate that Type II cells are activated following NM exposure; thus, pro-SP-C, a marker of activated Type II cells, was up-regulated, along with iNOS and COX-2. Type II cell production of iNOS and COX-2-derived mediators has been implicated in inflammation induced by LPS, bleomycin and ozone (Card et al., 2007; Cheng et al., 2016; Punjabi et al., 1994). It remains to be determined if Type II cells play a similar proinflammatory role in the lung following NM exposure. As observed in lung macrophages, in the absence of SP-D, expression of iNOS in Type II cells was increased. These findings are consistent with reports that Type II cells, like macrophages, rapidly up-regulate NF-κB following lung injury (Sunil et al., 2002). Our findings that loss of SP-D had no effect on NM-induced expression of COX-2 or pro-SP-C, indicates that these proteins are regulated by alternative signaling pathways in Type II cells. We also found that alveolar macrophages from WT and SP-D^−/− mice expressed relatively low levels of pro-SP-C. This is most likely due to a combination of phagocytosis of injured Type II cells and altered surfactant metabolism (Ikegami et al., 1998; Lopez-Rodriguez et al., 2017; Olmeda et al., 2017; Pinto et al., 1995).

Both monocytic and granulocytic MDSC were identified in the lungs of SP-D^−/− mice, but not WT mice. MDSC have been shown to play a role in the resolution of acute lung inflammation induced by Pseudomonas and Mycobacterium infection (Obregon-Henao et al., 2013; Rieber et al., 2013). This is due, in part, to the release of IL-10, which suppresses IL-12 production and promotes M2 macrophage activation (Kolahian et al., 2016). Increases in MDSC in SP-D^−/− mice may represent a compensatory attempt to limit proinflammatory macrophage activation in the absence of this pulmonary collectin. As observed with pro- and anti-inflammatory macrophages, loss of SP-D had no effect on numbers of these cells accumulating in the lung in response to NM. Additional studies are required to determine if the activity of these cells is altered in SP-D^−/− mice.

We also found that resident alveolar macrophages, identified as CD11b⁻CD11c⁺F4/80⁺Ly6G⁻Ly6C^lo, increased in the lungs of both WT and SP-D^−/− mice after NM exposure; this response was significantly greater in SP-D^−/− mice. Resident macrophages are thought to be important in maintaining lung homeostasis (Hussell et al., 2014). They originate from embryonic progenitors and are mainly sustained by proliferative self-renewal (Ginhoux et al., 2014; Hashimoto et al., 2013). Our observation that resident alveolar macrophages increase in SP-D^−/− mice after NM, without a significant change in numbers of infiltrating CD11b⁺Ly6G⁻ monocytic cells, are consistent with this notion.

Increases in lung macrophage activation in SP-D^−/− mice were linked to exacerbated inflammation and edema in response to NM. Additionally, in NM-treated SP-D^−/− mice, but not WT mice, there was evidence of alveolar-bronchiolar epithelial and goblet cell hyperplasia, bronchiectasis, and emphysema, as well as interstitial fibroplasia and collagen deposition. BAL cell, protein and phospholipid content, markers of damage to the alveolar epithelium and derangements in lung lipids, were also significantly increased in SP-D^−/− mice relative to WT mice, after NM administration. Exaggerated responses of SP-D^−/− mice to NM are most likely a consequence of baseline pulmonary inflammation. Similar structural and biochemical alterations in the respiratory tract have been described in rats after exposure to NM; however, in rats, these changes occurred more rapidly and were more pronounced, when compared with mice (Malaviya et al., 2015b; Sunil et al., 2011, 2014). Differences in the sensitivity of rats and mice to NM may be due to their differing capacity to bioactivate this cytotoxic vesicant. As observed in SP-D^−/− mice treated with ozone or bleomycin (Aono et al., 2012; Groves et al., 2012), macrophages accumulating in the lungs of SP-D^−/− mice in response to NM were enlarged, vacuolated and foamy in appearance, relative to macrophages in lungs of WT mice. These findings are consistent with an activated macrophage phenotype.

Oxidative stress plays an important role in mustard-induced pulmonary toxicity (Malaviya et al., 2016b). In response to oxidative stress, cells upregulate HO-1, a phase II stress response enzyme with anti-oxidant and anti-inflammatory activity (Rahman et al., 2006). In line with NM-induced oxidative stress, we observed increases in HO-1 expression in alveolar macrophages in both WT and SP-D^−/− mice. These findings are in accord with earlier studies in rats treated with NM (Malaviya et al., 2012, 2015b; Sunil et al., 2014). Loss of SP-D resulted in an exacerbated oxidative stress response to NM. Thus, macrophages staining for HO-1 were more numerous and were enlarged and foamy in appearance. Similar increases in macrophage HO-1 expression have been observed in SP-D^−/− mice after radiation-induced lung injury, which is also associated with oxidative stress (Malaviya et al., 2015a). SP-D has been shown to protect macrophages from oxidative damage (Bridges et al., 2000). Upregulation of HO-1 in macrophages, as well as Type II cells in SP-D^−/− mice may reflect oxidative stress in the absence of this protective protein.

Previous studies have demonstrated that loss of SP-D is accompanied by alterations in pulmonary mechanics (Groves et al., 2012). Similarly, we found that lung volume was reduced in SP-D^−/− mice, relative to WT mice; additionally, tissue resistance was increased, while tissue damping was reduced. These data are in line with reports that inflammation contributes to reduced pulmonary function (Hakansson et al., 2012). Despite aberrant functional responses in SP-D^−/− mice, they were less sensitive to the adverse effects of NM on pulmonary mechanics; thus, NM-induced decrements in lung volume and increases in tissue resistance and elastance were blunted. These findings are most likely due to chronic pulmonary inflammation in SP-D^−/− mice, which reduces their ability to respond to further mechanical injury. Analogous decreases in responsiveness have been described in SP-D^−/− mice following exposure to ozone or allergens (Atochina et al., 2003; Groves et al., 2012; Takeda et al., 2003).

The present studies demonstrate that SP-D regulates both proinflammatory/cytotoxic M1 and anti-inflammatory/wound repair M2 macrophage activation after NM exposure, and that in the absence of SP-D, there is dysregulation in the inflammatory responses of both macrophage subpopulations, resulting in greater structural damage, inflammation, and oxidative stress. Similar anti-inflammatory and/or anti-fibrotic activity of SP-D has been described in mice following pulmonary exposure to ozone, bleomycin, or radiation (Aono et al., 2012; Groves et al., 2012; Malaviya et al., 2015a), as well as after ventilation-induced lung injury in premature lambs (Sato et al., 2010). The fact that SP-D regulates both proinflammatory/cytotoxic M1 and anti-inflammatory/wound repair M2 macrophages suggests that differences between WT and SP-D^−/− mice are due to both increased sensitivity of SP-D^−/− mice to NM, and an impairment in tissue recovery and injury resolution. Our findings provide novel mechanistic insights into pathways regulating macrophage activation following NM-induced injury, which may be useful in the development of targeted anti-inflammatory therapies for lung disease.

SUPPLEMENTARY DATA

Supplementary data are available at Toxicological Sciences online.

FUNDING

This work was supported by the National Institutes of Health [grant numbers AR055073, ES004738, HL086621, and ES005022].